3 pillars of open science

  1. three pillars of Open Science

  2. FAIR applied to data

  3. FAIR applied to software (as much as possible)

data and software management plans

  • information required by funding bodies (ANR, OpenAire)
  • first draft must be transmitted no later than 6 months after contract is signed
  • template available on DMP OPIDOR
  • Contact ARDoISE data hub (through DMP OPIDOR or direct mail to guichet-ardoise@groupes.renater.fr)

what is a paper published without access to the underlying data

remember the Lancetgate.

no reproductibility = no public trust in Science

first step: your paper must provide a DAS (Data availibily statement)

FAIR principles (2016)

1. How to make your data findable

2. How to make your data accessible

3. How to make your data interoperable

4. How to make your data reusable

Findability

Findability through unique and perenne identifiers

publication dataset software
doi, HAL-ID DOI, Ark SWHID

Do not trust data repositories that do not provide identifiers to datasets

Findability in FAIR principles

Speaking about repositories

  • avoid commercial ones
  • check licence and policy
  • curated by someone ?
  • located in Europe (GDPR) ?

COSO guide

References